Self-assembly based on the cyanuric acid-melamine lattice - American

May 7, 1990 - Self-Assembly Based on the Cyanuric Acid-Melamine. Lattice1. Christopher T. Seto2 and George M. Whitesides*. Department of Chemistry ...
0 downloads 0 Views 346KB Size
pj-:]

6409

J . Am. Chem. SOC. 1990, 112, 6409-641 1 Scheme I

P

0

H-atom donor

oTi(IV)

-

+ "C

R

R

1

OTi(IV)

Cp2TiCI

R

0 OH

H+

-'TO'

R

Ti(lV)'

2

Scheme 11.

Reduction and Deoxygenation of Adenosine Epoxide"

-

a (69)

3b R PhCO

(a) Cp,TiCl/cyclohexa- 1,4-diene/room temperature. (b) 2Cp2TiCl/roomtemperature; % yields shown in brackets. of 5:l) arising from the @-eliminationsof either the Ti-oxo or the OMe species, with the former predominating.

the reaction proceeds via carbanion-like int rm di tes.

Supplementary Material Available: Details of typical experimental procedures and IH NMR, HRMS, and/or elemental analysis of new compounds (7 pages). Ordering information is given on any current masthead page. 10

12 (12 %)

1 1 (65Yo)

Self-Assembly Based on the Cyanuric Acid-Melamine Lattice Christopher T. Seto2 and George M. Whitesides*

. ..

Department of Chemistry, Harvard University Cambridge, Massachusetts 021 38 Received May 7, 1990

OH 13 (78%)

14 (16 %)

One limitation of this reduction protocol is revealed in the attempted reduction of monosubstituted terminal epoxides like 1,2-epoxydecane where up to 33% deoxygenation is observed even in the presence of a IO-fold excess of 1,4-cyclohexadiene. This may be due to the accessibility of the sterically unencumbered secondary/primary radicals to the Ti3+species resulting in further electron transfer reduction and subsequent deoxygenation rather than H atom transfer from cyclohexa- 1 ,4-diene. Finally, this deoxygenation reaction appears to be mechanistically different from the stereospecific low-valent W-mediated reaction reported by Sharpless et a1.I' Addition of either cisor trans-5-decene oxide to excess Cp,TiCI afforded an identical 73:27 mixture of cis- and trans-5-decenes. The mechanistic details of this reaction remain largely unknown except that the @-eliminations of alkoxides and the need for 2 equiv of Ti3+suggest that

( I I I (a) Sharoless. K. B.: Umbreit. M. A.: Nieh. M. T.: Flood. T. C. J . See also: Umbreit, M. A,; Sharpless, K.

Am. Chem. Soc.'1972, 94, 6538. B. Org. Synth. 1981, 60. 29

0002-7863/90/15 12-6409$02.50/0

Reaction of cyanuric acid (CA) with melamine (M) forms a stable, insoluble I : 1 complex C A S M . ~We, and others, believe that CAeM has a local structure represented by the lattice l.4 We have started a program in the design and synthesis of three-dimensional supramolecular assemblies based on the hydrogenbonding pattern of CA-M. This program is an extension of the strategy of molecular self-assembly that has been highly successful in forming quasi-two-dimensional monolayers on solid supports.5 ( I ) Supported by the National Science Foundation (Grants CHE-88-12709 to G.M.W.and DMR 86-14003 to the Harvard University Materials Research

Laboratory).

(2) NIH Training Grant Predoctoral Fellow, 1987-1988. (3) Kazuo, M.; Hironobu, K.; Koichiso, Y . ;Junichi, N. Jpn. Kokui Tokkyo Koho 1919. 79, 588. (4) Ostragorich, G.; Bacaloglu, R. Timisoura, Studii Cercefuri Stiinf. Chim. 1962, 9, 273-289. Finkel'shtein, A. I.; Rukevich, 0. S . Zh. Prikl. Speckfrosk. 1983, 38, 327-330. The X-ray powder pattern of CAyM, is consistent with our proposed structure. Zerkowski, J.; Graham, R.; Whitesides, G. M., unpublished results. For the crystal structure of CA.M.3HCI. see: Wang, Y . ;Wei, B.; Wang, Q.J . Crysfallogr.Spectrosc. Res. 1990, 20, 79-84.

Q 1990 American Chemical Society

Communications to the Editor

6410 J . Am. Chem. SOC.,Vol. 112, No. 17, 1990

-11.

z

,

90

150

I

80

'

I

'

70

'

I

8

50

60

'

I

40

'

I

30

'

,

20

'

I

'

10

I

'

00

PPM

Figure 1. Spectra of ' H N M R titration of hubM, (500 MHz, IO mM in CDCI,) with R'CA. The peak assignments are shown at the top of the figure. Scheme I. Self-Assembly of hubM3 with R'CA To Give a Supramolecular 1:3 Complex"

1

We set as our first objective the preparation of a soluble, stable derivative of the cyclic hexamer CA3-M3(see the boldfaced section in 1). To promote the formation of this structure, we preorganized the three M units by connecting them covalently to a central "hub", using "spokes" conformationally compatible with the desired CA3.M3 structure.6 Here we report that the tris(me1amine) hubM3 (2) reacts with neohexyl cyanurate (R'CA, 3) in CHCI, and forms a well-defined 1:3 complex (hubM3),(R'CA), (Scheme

3 (R'CA) R

R

Wh I

8'

2 (hubM3)

11.7

We monitored the titration of hubM, (10 mM in CDCI,) with solid aliquots of R'CA by IH NMR spectroscopy (Figure 1). The spectrum of uncomplexed hubM3 (bottom trace) has resonances that are broadened by self-association and restricted rotation around the amide and RNH-triazine bonds. At intermediate points in the titration, the spectrum shows resonances for the (hubM,) I(R'CA), complex against a background of uncomplexed hubM3. The resonances for the complex are sharp because it is locked into a single conformation and exchange between the complex and hubM3 in solution is slow on the N M R time scale. As the titration proceeds, the intensities of the resonances for (hubM,),(R'CA), increase until the 1:3 stoichiometry is reached, and beyond this point there is no further change in the ~pectrum.8~~ The peak assignments for ( ~ U ~ M , ) ~ ( R ' C A are ) , shown at the top of Figure 1. Two features support the assigned structure: First, several methylene protons (g,g' and q,q') of hubM3 are diastereotopic in ( ~ U ~ M , ) ~ ( R ' C and A ) , thus appear as two separate resonances. Second, the two imide N-H protons (w and (5) Whitesides, G. M.; Laibinis, P. E.Lungmuir 1990,6, 87-96. Bain, C. D.; Whitesides, G. M. Angew. Chem. 1989, 101, 522-528. (6) The importance of preorganization is clearly illustrated in the work of Rebek, Cram, and others. Rebek, J., Jr. Angew. Chem., fnr. Ed. Engl. 1990, 29,245-255 and references therein. Bryant, J.; Ericson, J.; Cram, D.J . Am. Chem. SOC.1990, i 12, 1255-1 256. Hydrogen-bonding networks have been used in many imaginative approaches to the construction of structurally defined, noncovalent complexes. Etter, M. Acc. Chem. Res. 1990, 23, 120-126. (7) All new compounds gave satisfactory 'H NMR (500 MHz), ' C NMR (125 MHz). and mass spectra. (8) We have examined the complexation of hubM, with a number of other

cyanuric and barbituric acid derivatives. These compounds also seem to lead A ) ~ , some of to complexes analogous in structure to ( ~ U ~ M , ) ~ ( R ' C although them contain up to 10%of other hydrogen-bonded species. We believe these other hydrogen-bonded species to be conformational isomers of a symmetrical 1:3 complex (Scheme I) in which one of the M units has been flipped 180° with respect to the plane of the CA3.M3 ring. These isomers have a dissymmetric arrangement of the three "spokes". (9) R'CA alone has low solubility in CDCI,. Beyond the 1:3 stoichiometry the R'CA in excess of that required to form (hubM,),(R'CA), does not go into solution.

R' = CH2CH2C(CH3)3

R" = CH2CeH&(CH&

-Or

(hubMdi(R'C43

"The third spoke of (hubM,),(R'CA), has been drawn schematically for the sake of clarity. The groups R' and R" were chosen to add solubility to the complex and to provide centers easily analyzed by ' H N M R spectroscopy.

w') of R'CA are in different hydrogen-bonding environments and thus appear as separate resonances in the complex even though they are identical in uncomplexed R'CA. NOESY studies are also consistent with the proposed structure. We observed NOES between (i) the tert-butyl protons of R'CA (v in Scheme I) and the adjacent aromatic protons of hubM, (I and m) and (ii) the methylene and terf-butyl protons of R'CA (t and v) and the adjacent tert-butyl protons of hubM, (s). In the (hubM,),(R'CA), complex these groups are close together. Several other methods support the 1:3 stoichiometry, (hubM3),(R'CA),. Vapor pressure osmometry (VPO) indicated a MW of 2720'O (calculated for (hubM,),(R'CA), = 2733) over the concentration range 2-16 mM in CHCI, at 37 O C . Titration of hubM3 (0.1 mM in CH2C12) with R'CA monitored by UV spectroscopy indicated a 1:3 complex. The complex is also stable enough to withstand chromatography. Reverse-phase TLC (eluted with 5% 2-propanol in CH2CI2)gave three spots: a major spot ( I 0) The values of three separate determinations were MW = 2670, 2640, 2850.

J . Am. Chem. SOC.1990, 112, 6411-6413 for the complex ( ~ u ~ M ~ ) , ( R ' C(Rf0.35-0.45) A)~ and minor spots for uncomplexed hubM3 (RfO.l 54.30) and uncomplexed R'CA (Ry O.O)." These data establish the composition and structure of (hubM3),(R'CA),. Preorganization, Le., joining the three melamine units by using the 1,3,5-C6H3R3hub and spokes conformationally compatible with the desired structure (Scheme I), is an important factor contributing to the stability of the complex. Reaction of monomeric derivatives of M and CA in solution forms complexes having approximately the correct molecular weight for CA3.M3 by VPO,'* but crystallization of these complexes yields extended linear structure^.'^ The apparent solution molecular weight in these systems may represent a true cyclic hexamer CA3.M3 or an average of linear and cyclic oligomers.

Acknowledgment. N MR instrumentation was supported by National Science Foundation Grant CHE-84-10774. Mass spectra were obtained by Dr. Andrew Tyler (Harvard University Mass Spectrometry Facility) using instrumentation provided by JEOL (USA) Inc. We thank Professor Robert Cohen (MIT, Chemical Engineering) for the loan of the vapor pressure osmometer and for helpful discussions.

Scheme I'

x-x

for X-ray diffraction. (12) VPO indicated a MW of 2710 for a I:l mixture of N,N'-bis(rl-tertbutylpheny1)melamine and 3,3,3-triphenylpropyI cyanurate (calculated for (N,N'-bis(4-tert-butylphenyl)melamine),(3,3.3-triphenylpropyl cyanurate), = 2370) over the concentration range 6-20 mM in CHCl, at 37 OC. The MW estimated by VPO depends strongly on concentration at lower concentrations. (13) Lehn, J. M.; Mascal, M.; DeCian, A,; Fischer, J. J . Chem. Soc., Chem. Commun. 1990,479-481. Zerkowski, J.; Seto, C.; Whitesides, G.M., unpublished results.

Synthesis and X-ray Structure of a Simple Metallaoxetane. Metal-Based Selectivity in Oxidative Addition Andrei A. Zlota, Felix Frolow, and David Milstein* Department of Organic Chemistry The Weizmann institute of Science Rehovot 76100, Israel Received March 12, 1990 2-Metallaoxetanes are often invoked as crucial intermediates in metal-catalyzed oxidations,' including cytochrome P-450 and biomimetic olefin epoxidationsa2 Their intermediacy is also postulated in various metal-mediated reactions of epoxide^,^ in olefin metathesis: and in conversions of carbonyls to alkenes by (1) For example: (a) Sharpless, K. B.; Teranishi, A. Y.; BBckvall, J. E. J . Am. Chem. Soc. 1977,99,3120. (b) Schrijder. M.; Constable, E. C. J . Chem. Soc.. Chem. Commun. 1982, 734. (c) Sheldon, R. A.; Kcchi, J. K. MetalCatalyzed Oxidations of Organic Compounds; Academic Press: New York, 1981. (d) Jorgensen, K. A. Chem. Rev. 1989,89,431. (2) See, for example: (a) Collman, J. P.; Brauman, J. L.; Meunier, B.; Hayashi, T.; Kodadek, T.; Raybuck, S. A. J. Am. Chem. Soc. 1985,107,2000. (b) Collman, J. P.; Kodadek, T.; Raybuck, S . A.; Brauman, J. L.; Papazian, L. M. Ibid. 1985, 107,4343. (c) Collman, J. P.; Brauman, J. L.; Meunier, B.; Raybuck, S . A.; Kodadek, T. Proc. Natl. Acad. Sei. UXA. 1984,81,3245. Avaria-Neisser. G. E.; Fish, K. N.; Imachi, M.; Kuczkowski, (d) Groves. J. T.; R. L. J. Am. Chem. Soc. 1986, 108, 3837. (e) Groves, J. T.;Watanabe, Y. Ibid. 1986, 108,507. (f) Rappe, A. K.; Goddard, W. A. Ibid. 1982, 104, 3287. (3) (a) De Pasquale, R. J. J . Chem. Soc., Chem. Commun. 1!?73, 157. (b) Blum, J. J. Org. Chem. 1977,42, 2299. (c) Blum, Milstein, D.; Buchman, 0.; J.; Zinger, B.; Milstein, D.; Buchman, 0.J. Org. Chem. 1978, 43, 2961. (d) Kamiya, Y.; Kawato, K.; Ohta, H. Chem. Lett. 1980, 1549. (e) Schwartz, J.; Hayasi, Y. Inorg. Chem. 1981.20, 3473. (f) Suzuki, M.; Watanabe, A.; Noyori, R. J. Am. Chem. Soc. 1980, 102, 2095. (8) Milstein, D.; Calabrese, J. C. J . Am. Chem. SOC.1982, 104, 3773. (h) Osborne, R. B.; Ibers, J. A. J . Organomet. Chem. 1982,232,273; (i) /bid. 1982,232,371. (j)Alper, H.; Arzoumanian, H.; Petrignani, J.-F.; Saldana-Maldonado, M. J. Chem. Soc., Chem. Commun. 1985, 340. (k) Miyashita, A.; Ishida, J.; Nohira, H. Tetrahedron Lett. 1986,27,2127. (I) Hase, T.; Miyashita. A.; Nohira, H. Chem. Lett. 1988,219. (m) Aye, K. T.; Ferguson, G.; Lough, A. J.; Puddephatt, R. J. Angew. Chem., Int. Ed. Engl. 1989, 28, 767.

0002-7863/90/ 15 12-641 1 S02.50/0

5

6

0-H l a (X=Br) l b (X=CI)

'L = PMe,; (1 1) We have not obtained crystals of the 1:3 complex that are suitable

641 1

C8Hl, = cyclooctene.

metal alkylidene~.~However, such complexes are exceedingly scarce and have been isolated in special cases where the metallaoxetane ring is stabilized by multiple cyano substitution6 or by an exocyclic double bond.' Only recently has the first example of a simply substituted 2-metallaoxetane been reported.* We report here (a) a straightforward route for the preparation of a simple metallaoxetane; (b) the crystallographic characterization of such a complex; (c) the structure of an analogous @-hydroxyalkyl complex, allowing direct evaluation of the structural consequences of ring closure; and (d) unusual metal-based chemoselectivity in oxidative addition of a bifunctional substrate. We had reported that epoxides undergo C-O oxidative addition to electron-rich rhodium and iridium complexes to yield 8-oxoalkyl metal hydrides, possibly by @-hydrideelimination of an intermediate metalla~xetane.~e.~However, attempts to isolate the primary oxidative addition product by use of a geminally disubstituted epoxide were not successful. Hence we planned the preparation of a metallaoxetane indirectly from the corresponding halohydrin (Scheme I). Addition of an equimolar amount of a 0.14 M toluene solution of the halohydrins 1 to a 0.04 M solution of the electron-rich iridium complex 2 in toluene under nitrogen at -30 OC leads after 30 min to formation of the hydrides 31° (ca. 70% yield), possibly by 0-H rather than C-X oxidative addition, followed by epoxide elimination. We find this surprising, since oxidative addition of alkyl halides to low valent metal complexes is usually a facile pr0cesslii taking place even in alcoholic solvents,I2 whereas such (4) Ivin, K. J.; Reddy, B. S. R.; Rooney, J. J. J . Chem. Soc., Chem. Commun. 1981, 1062. (5) (a) Tebbe, F. N.; Parshall, G.W.; Reddy, G. S . J . Am. Chem. Soc. 1978, 100, 361 1. (b) Buchwald, S. L.; Grubbs, R. H. /bid. 1983, 105, 5490. (c) Pine, S. H.; Zahler, R.; Evans, D. A.; Grubbs, R. H. Ibid. 1980,102,3270. (6) (a) Schlodder, R.; Ibers, J. A.; Lenarda, M.; Graziani, M. J . Am. Chem. Soc. 1974, 96,6893. (b) Lenarda, M.; Pahor, N. B.; Calligaris, M.; Graziani, M.; Randaccio, L. J . Chem. Soc., Dalton Trans. 1978, 279. (c) Lenarda, M.; Ros, R.; Traverso, 0.; Pitts, W. D.; Baddley, W. H.; Graziani, M.Inorg. Chem. 1977, 16, 3178. (7) Ho, S . C.; Hentges, S.; Grubbs, R. H. Organometallics 1988, 7,780. (8) Klein, P. D.; Hayes, J. C.; Bergman, R. G.J . Am. Chem. SOC.1988, I IO, 3704. (9) Milstein, D. J . Am. Chem. Soc. 1982, 104, 5227. ; NMR (pyridine-d5) (10) 3e: IR (Nujol) 2161 (m, +H), 951 (vs, u ~ ) 'H 6 1.49 (t, J+H = 3.7 Hz, 18 H, 2PMe,), 1.20 (d, J = 10 Hz, 9 H, PMe3), -21.6 (dt, JPI-H= 19.4 Hz, J ~ - = H 14 Hz, 1 H); plP('H)NMR (pyridine-d5) 6 -40 20 Hz, 2 P), -44 (t. Jp-pI = 20 Hz, 1 P). (d. J+pt (l!) (a) Collman, J. P.; Hegedus, L. S.;Norton, J. R.; Finke, R. G. Principles and Applications of Orgatwtransition Metal Chemistry; University Science Books: Mill Valley, CA, 1987; pp 284-321. (b) Stille, J. K.; Lau, K. S. Y. Ace. Chem. Res. 1977, IO, 434. (12) Oxidative addition of organic halides in the prsence of alcohols is a key step in metal-catalyzed carbalkoxylation of these substrates. See, for example: Milstein, D. Ace. Chem. Res. 1988, 21,428 and references therein.

0 1990 American Chemical Society